Chowhound's moderators and I did this same work day in and day out for the better part of a decade. It wasn't an automated process; we relied on deep experience hosting a forum that many characters found attractive for marketing purposes as a way to guerrilla market and foment word-of-mouth via fake testimonials. If the operation was going to have any value, we needed to weed that stuff. So we got good at it. Work hard/long enough on something - even something seemingly impossible - and you'll get good at it.In addition to grading reliability, Fakespot readjusts the Amazon rating, disregarding the phony ratings and re-averaging the remainder. Here's an example, where Fakespot adjusted Amazon's 4.5 star rating down to 2.5 stars:
But there's a mystery: very often Fakespot's adjusted rating is exactly the same....even though the product was heavily fake-rated. Two examples:
How can a rating be fraudulent yet also accurate?
It's a deep question. I suspect that even Fakespot can't explain it. But I can, after spending a decade immersed in this extremely thin slice of human experience.
I'll give the answer tomorrow [here it is]. Feel free to guess in the comments.
No comments:
Post a Comment